EO4Wetlands

The restoration of wetlands, critical ecosystems, is vital in view of the challenges of climate change. The full-scale experiment of depoldering offers a unique opportunity for EO4Wetlands to monitor this restoration using the complementarity between different wavelengths and scales of observation (in situ, drone and satellite).

[fr]

Satellite view and drone orthoimage overlay of the LLHPP site

Earth observation for multi-scale monitoring of wetlands restoration

Overview

Wetlands are among the critical ecosystems that play a major role in climate change, biodiversity and hydrology. The 2017 World Wetlands Days particularly highlighted how wetlands help cope with disasters such as floods and tides. As part of the Polder2C's Interreg project, experiments on the polder of the Living Lab Hedwige-Prosperpolder (LLHPP), on the Dutch-Belgian border, will lead to the partial destruction of the site and to wetland restoration. This is a unique opportunity to monitor its evolution in a maritime context of rising sea levels and this is the main objective of the SCO EO4Wetlands project.

Methodology and role of satellite data

The development of Earth observation data offers new tools to achieve this objective. In this project, data from the Sentinel-1 (radar images), Sentinel-2 (multispectral images), Sentinel-3 (thermal infrared images) and Landsat and their complementarities, will be used to study and better understand the evolution and roles of wetlands.

This variety of satellite data allowstra to better estimate

their biomass,
ltheir biogeochemical functions and
the water content in the vegetation and the critical zone.

NB: The Critical Zone refers to the very thin surface area of our planet, which is located at the interface between the atmosphere and the continental crust. It includes a large number of components often studied by different disciplines but all interconnected: weathered rocks, deep waters, surface waters, soils, aerial ecosystems and subterranean microbiota, lower atmosphere (OZCAR).

This information is based on

classification of visible and near-infrared data via NDVI (Normalized Difference Vegetation Index) and/or NWDI (Normalized Difference Water Index);
monitoring of wetland dynamics due to variations in content surface water the with, for example, radar data;
implementation of thermal infrared data in the existing classification as surface temperature changes leading to the estimation of daily wetland dynamics.

In detail, Sentinel-2 images will allow the development of a land-use based classification method to monitor the evolution of vegetation through time and space following the destruction of the dam. The new understanding of the temporal and spatial evolution of the vegetation and its biogeochemical functions will provide information that can be applied to other wetlands around the world. The high revisit time over the polder by the SLSTR (Sea and Land Surface Temperature Radiometer) sensor on board Sentinel-3A and 3B (day and night) will allow the study of the diurnal evolution of the surface temperature. This will provide indirect information on surface soil moisture. In addition, during flooding and/or sea surge events, the spatial evolution of the sea front can be delineated using Sentinel-3 data or using a combination of Sentinel-1 and -3 in case of cloudy conditions.

The spatial evolution of vegetation and surface temperature will allow to investigate the advantages and disadvantages of methodologies combining satellite data with very high spatial resolution images obtained with UAVs in the visible and thermal infrared spectra. In order to bridge the gap between UAV and satellite scales, we will rely on 1) in-situ recording of moisture content and surface temperature at specific locations in the LLHPP polder, in order to calibrate the satellite data and 2) the use of ECOSTRESS on board the ISS (70 m spatial resolution) which could play a bridging role between Sentinel-3 and the UAV survey in the thermal infrared wavelength bands.

Finally, the project aims to provide a wetland monitoring tool, which will bring together all wetland data obtained from in-situ sensors, UAVs and satellites, for end-users. The operational output, based on automatic recognition, will be a synoptic view of surface condition changes at different spatial and temporal scales. The tool will thus help end-users in charge of flood defence systems to monitor changes in vegetation and the potential flood front in case of emergency. In addition, as older satellite databases are available (since ~2016 for Sentinels and the 1990s for Landsat), historical studies can be undertaken taking into account the past and future impact of climate change on wetlands.

Application site(s)

Living Lab Hedwige-Prosperpolder, Belgium/Netherlands

Data

Satellite	Sensors	Spatial resolution (m)	Revisit time (days)
Sentinel-1	SAR (C band)	5	~6
Sentinel-2	Multispectral	~10	~5
Sentinel-3	Thermal infrared (TIR)	1000	1
Landsat	Multispectral + TIR	90	16
ECOSTRESS	TIR	70	1
Pleiades	Visible	0.5	on demand/archive

Results – Final products

Geomatys will develop the operational tool using its Examind software suite. Both Examind-Community and Examind-Datacube will be provided to meet the needs of this project. Based on a pre-existing classification algorithm developed by Cerema (french agency for ecological transition), the tool will be implemented using Examind-Datacube and the results will be disseminated using Examind-Community. This software stack is provided under an Apache2 Open Source license, and will be used to create a web portal that will provide dedicated tools to serve the needs of end-users.

The results will be :

the production of maps of the evolution of the vegetation over time in the polder (time series classification method);
assessment of the soil water content in the critical zone by recording the diurnal evolution of the surface temperature;
determination of the water front, especially during emergency events.

All the maps will provide a digital elevation model made locally by drone and complete at the polder scale using Pleiades data. In addition, some in situ sensors will record soil water content and temperature at specific locations in the LLHPP. These measurements will be integrated live into the tool through wireless communication and the Open Geospatial Consortium (OGC) sensorThings service hosted by Examind. These data will be taken into account for the calibration of satellite data.

This solution will be hosted on a Data Information Access Service (DIAS) infrastructure.

References

EO4Wetlands on GEO Knowledge Hub, a long lasting digital repository created by the Group on Earth Observations: https://doi.org/10.60566/wkbrj-ngd08

Gayet, G. et al. Guide de la méthode nationale d’évaluation des fonctions des zones humides–version 1.0. Onema, collection Guides et protocoles(2016).
Bhatnagar, S. et al. Mapping vegetation communities inside wetlands using Sentinel-2 imagery in Ireland. International Journal of Applied Earth Observation and Geoinformation 88, 102083 (2020).
Adam, E., Mutanga, O. & Rugege, D. Multispectral and hyperspectral remote sensing for identification and mapping of wetland vegetation: a review. Wetlands Ecol Manage 18, 281–296 (2010).
Guo, M., Li, J., Sheng, C., Xu, J. & Wu, L. A Review of Wetland Remote Sensing. Sensors 17, 777 (2017).
Whyte, A., Ferentinos, K. P. & Petropoulos, G. P. A new synergistic approach for monitoring wetlands using Sentinels -1 and 2 data with object-based machine learning algorithms. Environmental Modelling & Software 104, 40–54 (2018).
Rapinel, S., Clément, B. & Hubert-Moy, L. Cartographie des zones humides par télédétection : approche multi-scalaire pour une planification environnementale. Cybergeo : European Journal of Geography (2019) doi:10.4000/cybergeo.31606.
Sánchez-Espinosa, A. & Schröder, C. Land use and land cover mapping in wetlands one step closer to the ground: Sentinel-2 versus landsat 8. Journal of Environmental Management 247, 484–498 (2019).
Hubert-Moy, L., Clément, B., Lennon, M., Houet, T. & Lefeuvre, E. Etude de zones humides de fond de vallées à partir d’images hyperspectrales CASI : Application à un bassin versant de la région de Pleine-Fougères (Bretagne, France). Photo-Interprétation. European Journal of Applied Remote Sensing 39, 33–43 (2003).
Amani, M., Salehi, B., Mahdavi, S. & Brisco, B. Spectral analysis of wetlands using multi-source optical satellite imagery. ISPRS Journal of Photogrammetry and Remote Sensing 144, 119–136 (2018).
Muro, J. et al. Land surface temperature trends as indicator of land use changes in wetlands. International Journal of Applied Earth Observation and Geoinformation 70, 62–71 (2018).
Muro, J. et al. Short-Term Change Detection in Wetlands Using Sentinel-1 Time Series. Remote Sensing 8, 795 (2016).
Antoine, R. et al. Electric potential anomaly induced by humid air convection within Piton de La Fournaise volcano, La Réunion Island. Geothermics65, 81–98 (2017).
Antoine, R. et al. Thermal infrared image analysis of a quiescent cone on Piton de la Fournaise volcano: Evidence of convective air flow within an unconsolidated soil. Journal of Volcanology and Geothermal Research 183, 228–244 (2009).
Lopez, T. et al. Subsurface Hydrology of the Lake Chad Basin from Convection Modelling and Observations. Surv Geophys 37, 471–502 (2016).
Lopez, T. et al. Thermal anomalies on pit craters and sinuous rilles of Arsia Mons: Possible signatures of atmospheric gas circulation in the volcano. Journal of Geophysical Research: Planets 117, (2012).